Research Article

RNA stabilization by a poly(A) tail 3′-end binding pocket and other modes of poly(A)-RNA interaction

See allHide authors and affiliations

Science  05 Feb 2021:
Vol. 371, Issue 6529, eabe6523
DOI: 10.1126/science.abe6523

You are currently viewing the abstract.

View Full Text

Log in to view the full text

Log in through your institution

Log in through your institution

Poly(A) and RNA: More ways to interact

One of the most conserved RNA modifications across organisms is the addition of a polyadenosine, or poly(A), tail to the 3′ end of RNA molecules. Cis-acting RNA stabilization elements, such as the triplex-forming element for nuclear expression (ENE), retard RNA decay, thereby controlling the maturation and abundance of cellular transcripts. Torabi et al. determined the high-resolution crystal structure of a double ENE complexed with 28-mer poly(A) to reveal new modes of the RNA-RNA interaction, including a pocket motif that protects the extreme 3′ end of the poly(A) tail. Discovery of such interactions opens new venues to better understanding poly(A) tail function in RNA biology.

Science, this issue p. eabe6523

Structured Abstract


The polyadenylate [poly(A)] tail is one of the most conserved modifications of RNA molecules and is an important contributor to RNA function. In conjunction with other cis-acting RNA elements, the poly(A) tail plays a central role in RNA stabilization. The element for nuclear expression (ENE), which contains a U-rich internal loop (URIL) flanked by short double helices, stabilizes RNA by sequestration of the poly(A) tail via formation of a triple helix. ENE elements have been identified in evolutionarily diverse cellular and viral RNAs. Double-domain ENEs (dENEs), containing two URILs separated by a predicted double-helical region, appear in transcripts produced by plant and fungal transposons. All dENEs exhibit several distinguishing sequence features: three highly conserved adenosines (the adenosine triad) and a striking pyrimidine/purine (Y/R) bias in the composition of the URIL-flanking stems. The structural importance of these features in poly(A) tail protection has remained unknown.


To understand the mechanism of poly(A) tail sequestration by dENEs, we solved the crystal structure of a rice transposase mRNA dENE, TWIFB1, complexed with a poly(A)28. Complementary cellular and biochemical approaches then validated the contributions of the key structural elements to dENE function and provided mechanistic insights into poly(A) tail 3′-end protection.


The structure of the dENE+poly(A)28 complex at 2.89 Å resolution revealed multiple modes of interaction between the dENE and poly(A), including (i) two predicted canonical major-groove triple helices, (ii) a previously unnamed subclass of A-minor interactions between poly(A) and RNA double helices, (iii) an unrecognized quintuple-base motif that transitions poly(A) from minor-groove interactions to the major-groove triplexes, and (iv) a novel poly(A) 3′-end binding pocket. The 5′ region of the poly(A) interacts with the upper dENE domain while its extreme 3′ end is buried in the pocket motif of the lower domain. Poly(A) adenosines are inserted into the minor groove of the upper Y/R-biased stem through their Watson-Crick and/or Hoogsteen edges, forming what we call a WC/H A-minor motif, followed by the quintuple-base transition motif and subsequently a major-groove triple helix composed of four U-A•U triples. Similar structural features engage the poly(A) in the lower dENE domain but, unlike the upper dENE domain in which the 3′ end of poly(A) can freely exit the upper major-groove triple helix, they are followed by interactions with the pocket motif. The pocket, which is formed through stacking of the bases of the adenosine triad (cyan in the figure) within the major groove of the lower dENE stem, engulfs the 3′-most adenosine of the poly(A) and poises its 3′-OH group to form a hydrogen bond with a phosphate in the backbone of the pocket-forming RNA residues. Cell-based and biochemical assays confirmed the contributions of these structural features to poly(A) binding and demonstrated that the pocket motif augments sequestration of the poly(A) tail by protecting its 3′-most adenosine through a steric mechanism. The WC/H A-minor and quintuple-base transition motifs are found in other RNA structures.


Structural features uncovered in the dENE+poly(A)28 complex underscore the ability of the poly(A) tail to form extensive and underappreciated RNA-RNA interactions. Several consecutive adenosines can form a WC/H A-minor motif by interacting with a Y/R-biased stem, possibly contributing to the stabilization of polyadenylated RNAs. Moreover, a poly(A) 3′-end binding pocket composed solely of RNA provides insight into how polyadenylation can protect an RNA’s extreme 3′ end. Our data raise the possibility that comparable RNA-poly(A) interactions exist in other RNAs and play more pervasive roles in RNA biology than are currently known.

Structural overview of the TWIFB1 dENE+poly(A) complex.

(A and B) Multiple modes of interaction between the dENE and poly(A) tail are shown in cartoon representation (A) based on the crystal structure of the dENE (green) + poly(A)28 (purple) complex (B). The poly(A) 3′-end binding pocket formed by the stacked bases of the adenosine triad (cyan) lies in the major groove of the lower dENE stem. Non-native nucleotides in the crystallization construct are represented by a thick gray line. (C) Surface representation showing how the pocket motif engulfs the 3′-most adenosine of poly(A), as well as the hydrogen bond (dotted line) between the poly(A) 3′-OH group and a backbone phosphate of the pocket motif.


Polyadenylate [poly(A)] tail addition to the 3′ end of a wide range of RNAs is a highly conserved modification that plays a central role in cellular RNA function. Elements for nuclear expression (ENEs) are cis-acting RNA elements that stabilize poly(A) tails by sequestering them in RNA triplex structures. A crystal structure of a double ENE from a rice hAT transposon messenger RNA complexed with poly(A)28 at a resolution of 2.89 angstroms reveals multiple modes of interaction with poly(A), including major-groove triple helices, extended minor-groove interactions with RNA double helices, a quintuple-base motif that transitions poly(A) from minor-groove associations to major-groove triple helices, and a poly(A) 3′-end binding pocket. Our findings both expand the repertoire of motifs involved in long-range RNA interactions and provide insights into how polyadenylation can protect an RNA’s extreme 3′ end.

View Full Text

Stay Connected to Science